Throughput Analysis for Full-duplex Sensing in Non-time-slotted Cognitive Radio Network
Jingwen Zhang, Fanggang Wang, Zhangdui D. Zhong · 2018
In this paper, a non-time-slotted cognitive radio network (CRN) is considered, where the primary user (PU) can change its state between transmission and mute at any time within one secondary user (SU) frame. In such a case, the traditional half-duplex sensing, which divides the secondary transmission process into the sensing and transmission stages, has a problem that fails to detect the PU transition during the transmission stage. To overcome this problem, the full-duplex sensing is adopted, which allows the SU to sense the spectrum and transmit signals simultaneously. The SU throughput per frame is evaluated to determine the optimal duration of the SU frame. Note that the throughput scales with the duration of the SU frame in the absence of the PU. However, the PU may start to transmit at any time within the SU frame. Hence, we formulate a throughput optimization problem with respect to the duration of the SU frame, under the constraint that the interference induced by the SU does not exceed the interference temperature of the PU, which is nonconvex. We further indicate that the optimal solution can be achieved when considering some practical constraints.